Variability of somatosensory evoked potential and motor evoked potential change criteria in thoracic spinal decompression surgery based on preoperative motor status

Abstract

Background context: Combined somatosensory- and motor-evoked potential (SSEP and MEP) changes for predicting prognosis in thoracic spinal surgery have been variably reported.

Purpose: We aimed to explore the validity of combined SSEP and MEP for predicting postoperative motor deficits (PMDs) in thoracic spinal decompression surgery (TSDS) and identify a relatively optimal neurophysiological predictor of PMDs in patients based on preoperative motor status.

Study setting: Retrospective study.

Patient sample: A total of 475 patients were analyzed.

Outcome measures: A reduction in muscle strength by more than or equal to one manual muscle testing (MMT) grade postoperatively compared with the preoperative MMT grade was identified as PMDs. Postoperative motor deficits were detected by comparing the preoperative and postoperative physical examination findings in short- and long-term follow-up visits.

Methods: All patients were divided into two subgroups according to preoperative motor status. The following data were collected: (1) demographic data; (2) IONM (intraoperative neuromonitoring) data; and (3) postoperative motor outcomes. Binary logistic regression analysis was performed to assess the efficacy of IONM change to predict PMDs. A receiver operating characteristic curve (ROC) was used to establish optimal IONM warning criteria.

Results: Ninety-eight patients had severe preoperative motor deficits (Group S), and 377 patients did not (Group N). MEP and SSEP change was effective for predicting PMDs in the short term (p<.01) and long term (p<.01) for TSDS patients. In Group N, the cutoff values for predicting PMDs in the short term were a decrease of 65% in SSEP amplitude and 89.5% in MEP amplitude of the baseline value. Furthermore, the cutoff values for predicting PMDs in the short term were durations of change of 24.5 minutes for SSEP and 32.5 minutes for MEP. In Group S, however, the cutoff values for predicting PMDs in the short term were a decrease of 36.5% in SSEP amplitude and 59.5% in MEP amplitude of the baseline value. Moreover, the critical values for predicting short-term PMDs were durations of change of 16.5 minutes for SSEP and 17.5 minutes for MEP.

Conclusions: The optimal IONM changes for prediction vary depending on preoperative motor status. Combined SSEP and MEP are excellent for predicting PMDs in TSDS.

Keywords: Intraoperative neuromonitoring; Motor-evoked potential; Postoperative motor deficits; Somatosensory-evoked potential; Thoracic spinal decompression surgery.